Strained ring systems. XI. Synthesis of benzobicyclo[2.2.0]hexa-2,5

May 1, 1972 - hex-2-ene, and benzobicyclo[2.2.0]hex-5-en-exo-2-o1. Richard N. McDonald, D. G. Frickey, Gary M. Muschik. J. Org. Chem. , 1972, 37 (9), ...
0 downloads 0 Views 620KB Size
1304 J . Org. Chem., VoZ. 37, No. 9,197W

MCDONALD, FRICKEY, AND MUSCHIK

by preparative glpc (10 ft X 0.25 in. 10% SE-30 a t 170"). The spectral properties of this material were identical in all respects to those of the material isolated from the acid-catalyzed rearrangement of tricyclic ketone 5 (vide supra). Acid-Catalyzed Rearrangement of Octalone 21.-To 821.4 mg of octalone 2111 in a 10-ml Erlenmeyer flask was added 2 ml of concentrated sulfuric acid. The flask was stoppered and the brown mixture was stirred for 2 days a t room temperature. The mixture was poured into 16 ml of ice-water and extracted with ether (2 X 16 ml). The combined extracts were washed with 20 ml of i N sodium hydroxide and dried over magnesium sulfate. The solvent was evaporated to afford 646.8 mg of yellow oil. Glpc analysis of the product (6 ft X 0.26 in. 10% FFAP a t 180') revealed the presence of two components in a ratio of 48:52.

The two components were identified as the a,P-unsaturated ketones 8 and 9, respectively, by comparison of their spectra with those of authentic samples.

Registry No.-5, 17159-66-9; 6 , 18503-74-7; 6 2,4DNP, 18503-75-8; 7, 33830-72-7; 7 2,4-DNP, 3383073-8; 9, 33835-42-6; 10, 33835-43-7; 11, 33835-44-8; 18, 2582647-3; 19, 33835-46-0; 20, 33835-47-1 ; 25, 33835-48-2; 26, 33835-49-3. Acknow~edgments,-we thank the National Science Foundation and the *lfred Foundation for financialsupport.

Strained Ring Systems. XI." The Synthesis of Benzobicyclo[2.2.0]hexa-2,5-diene, Benzobicyclo[2.2.O]hex-2-ene, and Benzobicyclo[2.2.0]hex-5-en-exo-2-o11b RICHARD N. n/ICDONALD,* D. G. FRICKEY,'~ AND GARYM. MUSCHIK'~ Department of Chemistry, Kansas Btate University, Manhattan, Kansas

66608

Received October 6, 1971 This paper describes the cycloaddition of benzyne with cis-3,4-dichlorocyclobutene to give ezoJcis-5,6-dichlorobenzobicyclo[2.2.0]hex-sene (4), the disodium-phenanthrene dechlorination of 4 to benzobicyclo[2.2.0]hexa-2,5-diene (2), the diimide reduction of 2 to benzobicyclo[2.2.0]hex-2-ene (3), and the hydroboration of 2 to benzobicyclo [2.2.0]hex-6-en-ezo-2-01 (8). Some of the spectral features of these compounds are discussed.

When Dewar in 18672 reported his results from the oxidation of phenol, he suggested the use of a model by which he could construct the various structural isomers of a given molecular formula. For the formula CaH6, one of the structures written was that of bicyclo [2.2.0]hexa-2J-diene (l), which has become known as "Dewar benzene." Nearly 100 years later, van Tamelen and Pappas3 reported the successful synthesis of 1. Since that report various syntheses and studies of the chemistry of derivatives of 1 have been reported. I n our continuing program of the chemistry of molecules incorporating the [2.2.0] system, we felt that it would be most useful to develop a synthesis of benzobicy clo [ 2.2 .OIhexa- 1,5-diene ("hemi Dewar naphthalene")4 ( 2 ) which might also be applicable to the preparation of 1-substituted derivatives of benzobicyclo[2.2.0]hex-2-ene (3). Since 2 should be convertible to 5-substituted derivatives of 3, this approach would make available this set of compounds for further study. Such a synthetic approach has been achieved and is the subject of this paper. Synthesis. -Conceptually, the approach was to prepare cis 5,6 -dichlorobenzobicyclo [2.2.0]hex-2 -ene (4) by the cycloaddition of benzyne and cis-3,4-

-

(1) (a) Paper X : R,N. McDonald and E. P. Lyznicki, J. Amer. Chem. SOC.,98, 5920 (1971). (b) P a r t of this research was communicated in R. N. McDonald and D. G. Frickey, ibzd., 90, 5315 (1968). ( 0 ) Taken from the M.S. thesis of D . G. Frickey, 1968, and the Ph.D. thesis of G. M. Muschik, 1972. Edznburgh, 84 (1866); (h) W. Baker, (2) (a) J. Dewar, Proc. Roy. SOC. Chem. Brit., 1, 191 (1965), discusses Dewar's paper, and some of the efforts to synthesize 1 and its various derivatives. (3) E. E. van Tamelen and 9. P. Pappas, J. Amer. Chem. Soc., 8 6 , 3297 (i96a). (4) We are aware of only two other literature entries into this family of compounds which incorporate two of the classically considered structures of benzene fused into a single molecule: one is a bridged "Dewar anthracene" reported by Applequist and 9earle,6Band the second is a report of the 12dimethyl and 1,2,3,4-tetramethyl derivatives of 2." T h e preparation of "hemi Dewar biphenyl," a compound in which the moieties are joined b u t not fused, has been described by B u r t and Pettit.' (5) (a) D. E. Applequist and R. Seerle, J. Amer. Chem. SOC.,8 6 , 1389 (1964); (b) D. T. Carty, Tetrahedron Lett., 4753 (1969). (6) G . D. Burt and R. Pettit, Chem. Commun., 517 (1965).

a"'qC1 m: r.AmONO

+

COjH

c1 5

,

THF, CH.Cl, A

4

dichlorocyclobutene7 (5) and then to seek methods for dechlorination of 4 to 2 . Two methods were carried out for the cycloaddition reaction. One involved the in situ generation of the benzyne precursor and benzyne itself;sa only O.lyo of 4 was obtained. The second method involved isolation of the benzyne precursor, benzenediazonium-2-carboxylate (6),8b and allowing it to decompose thermally in the presence of 5. Yields of 4 ranging from 2.4 to 11.3% were obtained depending on the ratio of 5 :6 used. The impurities in the crude reaction mixture appeared to be largely aromatic from the nmr spectrum; one of these was benzoic acid. Chromatography on basic, activity I alumina and elution with carbon tetrachloride gave quite pure 4 in the first few fractions. The infrared spectrum of 4 was fairly simple, indicating a high degree of symmetry in the tricyclic structure. The nmr spectrum (CCl,, internal TMS) of 4 exhibited absorptions centered at T 2.80 (m, 4), 5.55 (m, J = 0.9 He, 2), and 5.92 (m, J = 0.8 Hz) 2). The aromatic hydrogens were assigned to the finely split multiplet at 7 2.80 which is only 0.2 ppm lower field than the center of the aromatic proton multiplet of benzocyclobutene (T 3.01).9 Irrespective of how we wish to rationalize the assignmentsI0 of the latter two absorptions (Cl, C4 bridgehead tu. Cg, Ce methine protons), the fact that the coupling constants are so small (7) M. Avram, I . Dinulescu, M . Elian, M. Farcasiu, E. Marica, G. Mateescu, and C. D. Nenitzescu, Chem. Ber., 97, 372 (1964). (8) (a) L. Friedman and F. & Logullo, 'I. J. Amer. Chem. SoC., 86, 1549 (1963): (b) M . Stiles, R. G. Miller, and U. Burkhardt, ibid., 86, 1792 (1963). (9) G. Fraenkel, Y . Asahi, M. J. Mitchell, and M. P. Cava, Tetrahedron, 90, 1179 (1965). (10) See S. J. Crista1 and G. W. Nachtigall, J. O r g . Chem., 82, 3738 (1967), for the assigned nmr spectrum of the related ezo,cis-5,6-dichlorobicyclo[2.2.1 Ihept-2-ene.

J . Org. Chem., Vol. 37, No. 9, 1972 1305

STRAINED RINGSYSTEMS requires the &,exo relationship of the chlorines in 4.1' The mass spectrum12of 4 was correct for a dichloride with the M + - , R/I+. 2, LI+. 4 peaks in a ratio of 1:0.625 : 0.25, which is well within experimental error of the expected ratio since the intensities of these peaks are very small. The base peak is m/e 128 and could correspond to the radical cations of naphthalene or "hemi Dewar naphthalene." Chemical proof of the gross structure of 4 was given by its dehydrohalogenation with potassium tertbutylate in tert-butyl alcohol to 2-chloronaphthalene in 45y0 yield. This product was characterized by comparison of its nmr spectrum with that of authentic material. The next step in this investigation was to examine methods for converting 4 into 2. Reaction of 4 with methyllithium L 3 did not take place and with sodium in l12-dimethoxyethane formation of only a small amount of 2 was observed. Treatment of 4 with lithium or sodium in ter.t-butyl alcohol and tetrah~drofuranl~ produced trace to small amounts of 2 along with varying amounts of tetralin, 1,4-dihydronaphthalene1 and naphthalene, and, in the case using lithium, a small amount of benzobicyclo [2.2.0]hex-2-ene (3).15 Finally, it was found that disodium-phenanthrene,16 a mild dehalogenating reagent, effected the dechlorination of 4 to 2 in 34% yie1d.I'

+

+

" L

The nmr spectrum of 2 (the samples always contain trace to small amounts of napht'halene) exhibited three finely split multiplets at T 2.95 (aromatic), 3.43 (olefinic), and 5.5'8 (bridgehead); t,his spectrum is reproduced in ref l b . The chemical shift of the olefinic protons is virtually identical with the values reported for the analogous prot'ons ( T 3.45) of bicyclo [2.2.0]hexa2,s-diene (1)3 and (7 3.43) benzobicyclo[2.2.l]hepta2,5-diene (7).1° The bridgehead protons of 2 absorbed 0.58 ppm downfield from those of l 3( T 6.16) and over 0.8 ppm downfield from those in 7 ( T 6.41, neat).I0 The ultraviolet spectrum of 2 was not sufficiently different from that of benzocyclobutenel* to warrant postulation of homoconjugation of the olefinic double bond with the aromat'ic ring As was previously mentioned, 2 rearranges to napht'halene. The rate constant for this rearrangement was determined in carbon tetrachloride at 38 f 0.2", the ambient probe temperature of the nmr spectrom(11) Cis vicinal coupling constants in four-membered rings are large (4-10 Hz) : see F. A. Bovey, "Nuclear Magnetic Resonance Spectroscopy," Academic Press, New York, N.Y . , 1969, p 362, (12) We thank Dr. R . W.Kiaer, University of Kentucky, for determination of this mass spectrum on a n RhZU-6E mass spectrometer. (13) G. Schroder and T. Martini, dngew. Chem., Int. E d . EngZ., 6, 806 (19671, reported a 70% yield of 2-chloro-3-fluorobieyc~o[2.2.@]hexa-2,5-diene from the reaction of 5,5-dichloro-6,6-difluorobicyclo[2.2.0lhex-2-ene and methyllithium. (14) P. G. Gassmaii and P . G. Pape, J . Org. Chem., 2 9 , 160 (1964). (15) The components were shown to be present from the nmr spectra of chromatographic fractions of the mixtuyes. (16) E. Vogle, H. Kiefer, and W. R. Roth, Angew. Chem., Int. Ed. Engl., 3 , 442 (19641, reported the use of this reagent in a debromination step to produce bicyclo (2.2.0 locta-2,4,7-triene. (17) This is a minimum yield. The yield of 2 used in a diimide reduction to 9 was 58% assuming a 90% yield in the reduction step. (18) M. P. Cava and D. R . Xapier, J . Amer. Chem. Soc., 80,2255 (1958).

eter. A half-life of 3.96 hr and a rate constant of sec-' were determined for this (4.85 h 0.48) X rearrangement under these conditions. Hydroboration of 2 followed by reaction of the intermediate borane with basic hydrogen peroxide produced benzobicyclo [2.2.0]hex-5-en-exo-2-01 (8) in 15% yield (based on 4). The ex0 configuration of the 2-hydroxy substituent in 8 is argued for by analogy with the specific exo hydroboration of bicyclo [2.2.0]he~-2-ene'~ and other bicyclic olefins, and the results of spin decoupling of the nmr spectrum of 8.20

H 8

To further characterize 2, it was reduced with diimide to 3.17 Attempted glpc of 3 led to rearrangement, with

3

l12-dihydronaphthalene and naphthalene as the only observed products. The mass spectrumzzof 3 showed a large >!I+ion (m/e 120, relative abundance 79%) with the h4+. - 1 peak as the base peak. This may involve cleavage of the C1-C4 bridge bond with loss of a hydrogen atom to produce benzylic cation 9. Further loss of a hydrogen atom then leads to m/e 128 (relative

+

9

10

abundance 72%), t'he naphthalene or 2 cation radical. The second major process appears to be the loss of CH3 to an ion m/e 115 (relative abundance 74%). A third path of electron impact fragmentation is the loss of ethylene from h i + *, the product of which may be 10 (m/e 102, relative abundance 10%). The nmr spectrum of 3 (reproduced in ref lb) consists of an AtB2 multiplet for the aromatic protons centered at T 2.93, a multiplet centered at T 6.19 for the bridgehead protons, and a complex multiplet between T 7.33 and 8.28 att'ributed to the methylene protons. The ultraviolet spectrum of 3 showed that the three large absorptions were shifted t'o somewhat lower wavelengths than t'hose of 2 and gave a reasonable correspondence to those of benzocyclobut'ene. a

(15) R. N. McDonald and C. E. Reineke, J. O P ~Ckern., . Sa, 1878 (1967). (20) Saturating in the methylene proton (xH3 and nHs) region ( r 7.6-7.85) the original methine (nHz) simplified to a broad einglet. The residual coupling was between HI end nHz and must he less than J = 1 Hz, which establishes the hydroxyl group as exo.11121 Saturation in the bridgehead proton (HLand Ha) region simplifies the methylene (xH3 and nHs) multiplet t o a complex doublet due to the remaining germinal coupling, Jnm J,H~,a chemical shift difference in xH8 and nHs, and cis ( J n ~ -% J n ~ s ) and trans ( J n ~ % - J x ~ 8 vicinal ) couplings. T h e triplet observed for nHz sharpened by saturating a t the bridgehead proton region. n and x refer t o endo and exo, respectively. (21) (a) K. B. Wiberg, V . Z. Williams, and L. E. Friedrich, J . Arne?. Chem. Soc.. 90, 5338 (1968); (h) K.B. Wiberg and D . E. Barth, ibid., 91, 5124 (1569). (22) We thank Dr. R . G. Cooks for determination of this mass spectrum on a n MS-9 mass spectrometer.

-

1306 J . Org. Chem., Vol. 37, No. 9, 1972

JICDONALD, FRICKEY, AND MUSCHIK

Experimental Sectionz3

mp 58'). The nmr spectrum of the product was identical with that of an authentic sample. Benzenediazonium-2-carboxylate.-This compound was preBenzobicyclo[2.2 .O] hexa-2,5-diene (2) .-To a magnetically pared according to the published procedure.8b Yields of 33-617, &red solution of 2.021 g (11.79 mmol) of phenanthrene in 10 of brown to light buff needles were obtained. The infrared ml of freshly distilled glyme (distilled from calcium hydride and spectrum compared well to that reported!b Caution: This then from lithium aluminum hydride) in a 25-m1, dry, two-necked compound is explosive when subjected to friction, especially flask fitted with a rubber septum and a gas inlet was added when scraped on hard surfaces. 0.700 g (30.4 mg-atoms) of sodium a t ice bath temperature. (The ezo,cis-5,6-Dichlorobenzobicyclo[2.2.0] hex-2-ene (&).-To a sodium was cut immediately prior to addition and added quickly in solution of 13.2 g (107 mmol) of cis-3,4-dichloro~yclobutene~ so as to present a clean surface t o the phenanthrene solution.) 100 ml of dioxane (distilled from sodium) was added 3.14 g Bfter a few minutes of stirring under nitrogen, the solution be(21.2 mmol) of freshly prepared benzendiazonium-2-carboxylate. came very dark green. Thirty minutes was allowed for the The resultant slurry was magnetically stirred a t 48-52' under reaction of the sodium with the phenanthrene. nitrogen until the mixture became homogeneous (3.25-4.5 hr The green solution was then recovered from the flask by means depending on the purity of the benzyne precursor). The soluof a syringe and charged into a solution of 0.782 g (3.93 mmol) tion was dark brown a t this point. The dioxane and the cisof 4 (75-90% pure) in 10 ml of anhydrous glyme under nitrogen 3,4-dichlorocyclobutene were removed under vacuum using a a t room temperature in a 50-ml two-necked flask fitted as above. 33-cm Vigreux column with a maximum pot temperature of 70", The first few drops were decolorized and a white precipitate leaving a dark brown viscous residue. This residue was disformed immediately. After 30 min of stirring under nitrogen, solved in about 100 ml of ether and filtered to give a colored the mixture became brown and never lightened much more during solution. The ether was evaporated in a molecular still, and the the 1.25-hr reaction time. product was sublimed at room temperature and 0.05 mm from The reaction mixture was dissolved in 100 ml of ether, which the residue as a white crystalline solid mixed with a light yellow was washed with four 50-ml protions of water and dried (MgS04) oil. The oil was separated from the solid by rubbing the mixin the freezer. Evaporation on a rotary evaorator to near dryture on a piece of absorbent paper. This afforded 0.48 g (11.3y0 ness gave a yellow, solid residue. Any residual ether or glyme yield) of crude product. Further purification was effected by was evaporated by means of an oil pump and collected in a Dry subliming half of the tots1 yield at room temperature and 0.05 Ice cooled trap. The product was distilled using a trap-to-trap mm, cleaning the cold finger and continuing the sublimation. apparatus at room temperature and 0.1-0.2 mm pressure and The second half of the sublimation was the purer sample (mp collected in a trap cooled in liquid nitrogen. The length of 70-79"), which was purified further by repeated sublimations. time of distillation was determined by stopping when a little The first half was impure and was purified by column chroof the starting material began to sublime onto the inside of the matography using neutral, activity I Woelm alumina with upper portion of the vacuum adaptor used for the distillation. carbon tetrachloride as an eluent with a fairly fast elution rate. The yield of the product as a clear, colorless liquid was 171 mg The first few fractions contained the purest product, mp 60-79', (347,). The unreacted starting material was recovered by subfrom the chromatography. This was purified further by reliming about one-fourth to one-third of the residual solid from peated sublimations. The melting point of the product was 79the distillation and chromatographing the sublimate on basic, 79.5" (sealed tube). The nmr, infrared, mass, and ultraviolet activity I alumina with carbon tetrachloride as the eluent. [263 nm (log e 3.19), 267 (shoulder) (3.23), 269 (3.42), and The nmr and ultraviolet [257.8 nm (shoulder) (log e 2.75), 263.4 275.3 (3.44) in cyclohexane] spectra all agreed with the assigned (2.97), 270 (3.12), 276.5 (3.15), 284.5 (2.37), 287 (2.32) in cyclostructure. hexane containing ea. 5.6% naphthalene impurity] spectra were Anal. Calcd for CloH&l%: C, 60.35; H , 4.02. Found: C, in agreement with the assigned structure of 2. 60.39; H , 4.14. Determination of the Kinetics of the Isomerization of 2 to I n several runs of this preparation, various ratios of the cisNaphthalene.-A sample of 5-10 mg of 2 was dissolved in 0.2 ml 3,4-dichlorocyclobutene to benzenediazonium-2-carboxylatewere of carbon tetrachloride (distilled from phosphorus pentoxide) used. Table I is a table of these ratios and their respective yields. to give a 0.17-0.39 1M solution. Using a capillary tube containing a 10% solution of methylene chloride in carbon tetrachloride as a standard, the amount of 2 present a t various times TABLE I was determined by integration. The nmr spectral absorption Yield of Cyclobutene: corresponding t o the T 3.42 absorption of 2 and the methylene diazonium salt 4, % chloride singlet were integrated and a ratio of the areas was ob7 . 3 3.7:l tained. The temperature a t which the study was done was 38 f 4.5 5.5:l 0.2", the ambient probe temperature. 10.4 8.O:l Benzobicyclo 12.2 .O]hex-2-ene (3).-A 250-mg (I.26 mmol) 10.0 4.0:l sample of dichloride 4 was dechlorinated t o 2 using the disodiumphenanthrene complex. Two trap-to-trap distillation fractions 11.3 5.0:1 were obtained which contain 2 with no apparent naphthalene 8.9 12.5:l present. These fractions were dissolved in 5 ml of dry methanol 2.4 1.5:l and combined with a yellow, methanolic slurry of dipotassium azodicarboxylate [prepared from 0.630 g (5.43 mmol) of azodiThese yields are based on the product sublimed from the crude carbonamideZ5]with stirring under nitrogen. To this was slowly as the product mixture with the benzenediazonium-2-carboxylate added a solution of 0.90 ml of glacial acetic acid dissolved in 5 limiting reagent as described in the above procedure. ml of dry methanol over a 2-hr period at room temperature. The Dehydrochlorination of e~o,cis-5,6-Dichlorobenzobicyc1o- yellow slurry changed to white after a portion of the acid had [2.2.0]hex-Z-ene.-4 (30 mg, 0.15 mmol) was dissolved in 4.5 ml been added. After stirring under nitrogen for an additional 5 of tert-butyl alcohol (distilled from sodium). To this was added hr, the reaction mixture was transferred t o a separatory funnel 24 mg (0.6 mg-atom) of potassium and the mixture was heated with 150 ml of ether. The ether layer was separated, washed under reflux overnight. Water (23 ml) was then added and the with eight 60-ml portions of water and 10 ml of saturated, mixture was extracted with three 25-ml portions of pentane. aqueous sodium bicarbonate solution, and dried (MgS04) in the The pentane extracts were washed with three 100-ml portions of refrigerator overnight. Evaporation of the ether left a yellow, water, dried (;CIgS04),and distilled to near dryness. The rest liquid residue which was trap-to-trap distilled to give 85 mg of 3, of the solvent was evaporated on a rotary evaporator to give 11 bp 25' (0.5-0.8 mm). This represents a 52% overall yield of 3 mg (457, yield) of crude 2-chloroiiaphthalene, mp 35-58' (lit.24 from 4. The nmr and ultraviolet [262 nm (OD,,, 0.94), 268 (1.17), 274 (1.28), and 285 (0.89)]z6 spectra were in agreement (23) All melting points were determined on a Kofler hot stage and are with the assigned structure.

-

corrected. Boiling points are uncorrected. Infrared and ultraviolet spectra were determined on a P-E 137 and Cary 11 spectrophotometer, respectively, while nmr spectra were obtained on a Varian A-60 spectrometer. Gas chromatographic analyses were performed using a F & M Model 500 temperature-programmed gas chromatograph. Microanalyses were done b y Galbraith Laboratories. (24) F. D. Chattaway and W. H. Lewis, J. C h e n . Soc., 875 (1894).

( 2 5 ) J , Thiele, Justus Lzebigs Ann. Chem., 271, 127 (1892). (26) Extinction coefficients were not determined for this spectrum due to an unknown degree of contamination by naphthalene and binder from tlc

from another preparation.

J . Org. Chem., Vol. 37, No. 9, 1972 1307

SUBSTITUTED TETRAHYDROAND HEXAHYDROFLUORENES Another preparation of 3 produced enough pure compound by column chromatography for elemental analysis. Anal. Calcd for CloHlo: C, 92.26; H, 7.74. Found: C, 92.09; H , 7.74. Benzobicyclo [ 2.2 .O] hex-5-en-em-2-01 (8).4 (150 mg) was converted to 2 as described above. Olefin 2 was immediately transferred to a reaction vessel containing 1 ml of tetrahydrofuran To this was added 0.45 ml of (distilled from LiA1H4)a t 5-10'. a 0.45 M diborane solution in tetrahydrofuran. After stirring for 1 hr, water was added dropwise to destroy excess dihorane followed by 0.1 ml of 3 N aqueous sodium hydroxide and 0.1 ml of 30y0 hydrogen peroxide. Stirring was continued for 1 hr. The reaction mixture wa6 extracted with 20 ml of ether, which was separated and washed with four 5-ml portions of saturated, aqueous brine. The organic layer was dried (MgSO4) a t O " , the ether was removed on a rotatory evaporator, and the liquid residue was chromatographed on activity 11, basic alumina to give 17

mg (15% based on 4 ) of 8: ir (neat) 3260,2950, 1460, 1060, and 745 cm-1; nmr (CC14,internal TMS) T 2.85 (m center, aromatic A2B2, 4), 5.75 (t center, C2H, l ) , 6.15 (m center, bridgehead H's, 2), 6.87 (9, OH, l ) , and 7.6-7.85 (m, CH2, 2). Anal. Calcd for ClaHlOO: C, 82.16; H, 6.90. Found: C, C, 81.98; H, 6.81.

Registry No.-2, 20847-82-9; 3, 20847-83-0; 4, 20902-25-4 ; 8,33905-59-8. Acknowledgment. -We wish to thank the National Science Foundation (GP-7818, GP-10691) for support of the research and the U. S. Department of Health, Education, and Welfare for an NDEA Title IV Fellowship to G. M. M. (1966-1969).

Synthesis of Angularly Substituted Tetrahydro- and Hexahydrofluorenesl HUGHW. THOMPSON* AND RICHARD E. NAIPAWER~ Department of Chemistry, Rutgers University, Newark,New Jersey 07108 Received June 16, 1971 A series of 7-methoxy-2,2-ethylenedioxy-l,2,3,9a-tetrahydrofluorenes has been synthesized in which the Sa substituent has been transformed from COOEt into CHO, CHNOH, COOLi, COONa, CONH2, CN, and COMe as well as into the previously reported groups COOMe, CH20H, and COOH. For each substituent the cis and trans compounds resulting from olefin reduction have been synthesized by routes which establish their stereochemistry. The influence of the stereochemistry at C-4a on the course of reactions at the group attached a t C-9a is discussed.

I n the hydrogenation of olefins over heterogeneous catalysts, the stereochemistry of reduction has been found to be influenced not only by the bulk of neighboring functiona,l groups but by attractive interactions between certain of these groups and the catalyst surface.3 These interactions, in contrast to steric effects, are responsible for addition of hydrogen cis with respect to the group involved. Our study of this phenomenon3 required us to synthesize a series of compounds lu, in which the angular group R represented a 4

5

MeOCH,CH,OH

Me0

R I O lU

HI

IC = cis; It = trans variety of common functional groups, as well as requiring us to prepare authentic samples of the corresponding cis (IC) and trans ( I t ) reduction products. We report here these syntheses, which were accomplished by transformations of the unsaturated carboethoxy compound (2u) and which illustrate reactivities and limitations in a system which is both severely crowded and acid sensitive. The results are instructive as to the relative usefulness in such an environment of a number of the existing methods for functional group transformation. The only cases for which we had already made entire (1) Abstractedin part from the P h . D . thesis of R . E. N. (2) NASA Predoctoral Trainee, 1966-1967. (3) H . W.Thompson, J . Org. Chem., 86, 2577 (1971).

sets consisting of all three compounds (u, c, and t) for a given functional group were carbomethoxy (3) and hydroxymethyl (4). Since the stereochemistry of each of these materials was unequivocally known, they were in practice the starting or reference points for our 'synthetic sequences. As these sequences in many instances parallelled each other for our three series of compounds (Schemes 1-111) we shall usually discuss the transformations in the unsaturated series as being typical of all three. We had synthesized the previously reported cis aldehyde (and established its stere~chemistry)~ by oxidation of the cis-hydroxymethyl compound 4c with CrOn in ~ y r i d i n e . ~The same procedure was successful for preparation of 5u and 5t; however, yields were consistently poor and we found that greatly improved yields for the entire aldehyde set could be obtained with the procedure employing dicyclohexylcarbodiimide and dimethyl sulfoxide.6 Treatment of the aldehyde 5u with hydroxylamine in ethanol proceeded smoothly to give the oxime 6u.' The unsaturated carboxylic acid 7u had been produced as previously described3 from the corresponding ester by saponification. The trans acid could also be obtained by saponification, albeit under more drastic conditions than were required for 2u, and the stereochemistry of 7t was thus related unequivocally t o the trans series. As none of the reduction methods tried on the unsaturated ester 3u gave appreciable quantities of cis ester, we could not prepare the cis acid by an analogous saponification. However, metal-ammonia reduction of the acid 7u provided 7c in good yield and (4) (5) (6) (7)

H. W .Thompson, i b i d . , 82, 3712 (1967). J. Schmidlin and A. M'ettstein, H e b . Chim. Acta, 46, 331 (1962). K. E. Pfitzner and J. G. Moffatt, J . Amer. Chem. Soc., 81, 5670 (1965). The same procedure was used t o obtain the cis and trans isomers.